EP0182134B1 - Method for operating a fail-safe multi-computer system with some not fail-safe input/output units - Google Patents

Method for operating a fail-safe multi-computer system with some not fail-safe input/output units Download PDF

Info

Publication number
EP0182134B1
EP0182134B1 EP85113416A EP85113416A EP0182134B1 EP 0182134 B1 EP0182134 B1 EP 0182134B1 EP 85113416 A EP85113416 A EP 85113416A EP 85113416 A EP85113416 A EP 85113416A EP 0182134 B1 EP0182134 B1 EP 0182134B1
Authority
EP
European Patent Office
Prior art keywords
data
input
output modules
telegrams
secure
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired - Lifetime
Application number
EP85113416A
Other languages
German (de)
French (fr)
Other versions
EP0182134A2 (en
EP0182134A3 (en
Inventor
Heinz-Jürgen Dr.-Ing. Lohmann
Rudolf Dipl.-Ing. Günther
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Siemens AG
Original Assignee
Siemens AG
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Siemens AG filed Critical Siemens AG
Priority to AT85113416T priority Critical patent/ATE72489T1/en
Publication of EP0182134A2 publication Critical patent/EP0182134A2/en
Publication of EP0182134A3 publication Critical patent/EP0182134A3/en
Application granted granted Critical
Publication of EP0182134B1 publication Critical patent/EP0182134B1/en
Anticipated expiration legal-status Critical
Expired - Lifetime legal-status Critical Current

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L1/00—Arrangements for detecting or preventing errors in the information received
    • H04L1/22—Arrangements for detecting or preventing errors in the information received using redundant apparatus to increase reliability
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00—Error detection; Error correction; Monitoring
    • G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
    • G06F11/0703—Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation
    • G06F11/0706—Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation the processing taking place on a specific hardware platform or in a specific software environment
    • G06F11/0721—Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation the processing taking place on a specific hardware platform or in a specific software environment within a central processing unit [CPU]
    • G06F11/0724—Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation the processing taking place on a specific hardware platform or in a specific software environment within a central processing unit [CPU] in a multiprocessor or a multi-core unit
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00—Error detection; Error correction; Monitoring
    • G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
    • G06F11/0703—Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation
    • G06F11/0751—Error or fault detection not based on redundancy
    • G06F11/0763—Error or fault detection not based on redundancy by bit configuration check, e.g. of formats or tags
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00—Error detection; Error correction; Monitoring
    • G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
    • G06F11/16—Error detection or correction of the data by redundancy in hardware
    • G06F11/20—Error detection or correction of the data by redundancy in hardware using active fault-masking, e.g. by switching out faulty elements or by switching in spare elements

Definitions

  • the invention relates to a method for operating a multi-computer system that is secure in terms of signaling technology, with at least two input / output modules that are not secure in terms of signaling technology, by means of which the computers of the multi-computer system use at least two-channel data telegrams from and / or to other computers and / or other data-receiving, outputting or processing systems , transmitted and via which they receive or deliver useful data secured by test data for the input / output modules from there in the form of telegrams with the same content.
  • Multi-computer systems For particularly security-relevant applications, such as railway safety, signal-safe multi-computer systems have been developed in which the data to be processed are treated independently of one another in several channels according to certain laws; Comparative procedures identify any data deviations in the individual channels and convert them into corresponding control commands, which act on the process to be controlled in the sense of a risk exclusion.
  • Multi-computer systems that are secure in terms of signaling not only have to be able to treat the data supplied to them in the specified manner, they also have to be able to determine whether they are not in terms of their signaling data present in safe input / output modules, for example as a result of a transmission fault or a component defect, are not falsified, in order to then prevent the evaluation of these falsified data.
  • a known method for securing the data transmission provides for the test data formed on the transmission side to be added to the actual user data, the corresponding test data to be reproduced on the receiving side according to the same laws on the receiving side and to be compared with the transmitted test data.
  • This type of data backup is generally more effective the more complex the test data becomes.
  • Unpredictable failure reactions of a module e.g. one equipped with highly integrated circuits
  • An input / output module can basically be identified by multiplying these modules and comparing their reactions with one another. For the present application, this means that the one or. Information to be output in each case must be written in parallel in at least two separate input / output modules and that the multi-computer system then checks the mutual agreement of the stored data. If an inequality is recognized when comparing the data stored in these modules, the computer system blocks the use of the stored information. This means that the first failure of an input / output module is absolutely harmless.
  • a defect in an input / output module should be recognized as such immediately and its effects rendered ineffective; However, such a defect should, if possible, not block the further data traffic of the computer system with the system elements which supply the input / output module affected by the defect with data or obtain data from there, and in particular should not lead to failure of the entire computer system.
  • a disadvantage of this design is the high effort for the input / output modules. This is particularly serious if the input / output modules are the interfaces to many other computers or other system elements with the same requirements. Then this additional effort has to be driven on all computers and this is very expensive.
  • One way of linking the two aforementioned backup methods to one another is to provide the data to be transmitted with test data and to evaluate them in a multi-channel manner.
  • This method is used in the circuit disclosed in DE-OS 26 09 107. After checking the redundancy in several mutually independent modules, the results formed by these modules are compared with one another. If one assumes that - as already explained - data falsifications in a register cannot be detected with absolute certainty by redundancy checks, then such falsifications cannot be reliably identified by a double redundancy check. This means that the circuit known from DE-OS 26 09 107 does not meet the high security requirements of signal-safe multi-computer systems.
  • the methods of the invention should be so transparent in themselves that it is possible to prove that data corruption has been safely accessed as a prerequisite for the approval of the method according to the invention in security-relevant applications.
  • Both methods are designed in such a way that they reliably react to data corruption in the input / output modules of a multicomputer system to be filled with matching data telegrams by preventing the evaluation of the faulty data and excluding the elements that are or possibly involved in the data corruption from further data transmission.
  • the further transmission of data takes place in the form of so-called long telegrams, each consisting of two telegrams of the same length, each with the same content, exclusively via those input / output modules and transmission channels that are not affected by the detected fault.
  • the two methods lead to different reactions in relation to the number of system elements affected by the respective disturbance:
  • the respective disturbance remains at least initially restricted to those system elements that were actually involved in a faulty data transmission, that is, to an off - And an input module of two system elements;
  • the transmission of data between all other system elements, including those affected by the malfunction and the other system elements, takes place initially - until any further malfunctions occur - using two-channel telegrams of normal length, then also in single-channel long telegrams for the system elements affected by these malfunctions .
  • the features of claim 2 provide to inform all system elements when a fault occurs and to ensure that all system elements in traffic with the system elements directly affected by the fault henceforth, for example, use single-channel using long telegrams.
  • the data transmission is permitted for a certain period of time after the finding of faulty data with additional data backup and allocation of a transmission channel not affected by the data corruption, which can be used to remedy the defect that has occurred.
  • the invention is based on exemplary embodiments shown schematically in the drawing explained in more detail.
  • FIGS. 1 and 2 show in FIGS. 1 and 2 a particularly advantageous structure of computer arrangements consisting of several multi-computer systems which are secure in terms of signaling and which react in different ways to incorrectly stored data
  • FIG. 3 is a functional diagram for explaining the methods according to the invention in a two-computer system which is secure in terms of signaling.
  • the computer arrangement shown in FIG. 1 consists, for example, of six multi-computer systems RS1 to RS6 which are secure in terms of signal technology and which are connected to the two buses B1 and B2 of a bus system via assigned input / output modules, via which the multi-computer systems communicate with one another in two channels. It is assumed that the computer system RS2 has to transmit data to the computer system RS1 in the form of a telegram.
  • the computer system RS2 activates its two input / output modules E / A21 and E / A22 for, for example, address-oriented transmission of data telegrams of the same content via the two buses B1 and B2 to the input / output modules E / A11 and E / A12 of the computer system RS1;
  • the data telegrams transmitted via the two buses can be displayed differently, for example, inversely.
  • the RS1 computer system that records the data checks the data stored in its input / output modules before they are recognized for compliance with the specified data format and for content-related correspondence.
  • the RS1 computer system may determine different data in its input / output modules E / A11 and E / A12.
  • the computer system then discards the transmitted data telegram and requests both Buses B1, B2 from the data transmitter, the secure multi-computer system RS2, send a new telegram transmission. If the data transmitted thereby satisfy the test conditions of the RS1 computer system, they are recognized, and the data sender can be acknowledged if the data are recognized. If the second attempt at data transmission also ends negatively, i.e. if the data recipient detects data in its input / output modules that differ in content or format, it will determine in a manner yet to be explained which of the data records stored in its input / output modules is faulty and which one Not. In the present example, the computer system RS1 may have recognized the data pending in its input / output module I / O11 as incorrect.
  • the two computer systems directly affected by the detected fault are also to communicate with other computer systems connected to the same bus system, it is advantageous to also inform these computers that certain input / output modules of the computer system which is detecting the fault and possibly the computer system, from which the corrupted data telegram originated for further data transmission are no longer available. This is to prevent a defect in a single input / output module from gradually blocking further input / output modules from other computer systems.
  • the multicomputer systems RS1 and RS2 which are directly affected by the disturbance correspond to one another and possibly with the multicomputer systems RS3 to RS6 which are not directly affected by the disturbance in a manner as will be explained in detail with reference to FIG. 3.
  • the transmitting multicomputer system in each case also checks the data pending in its input / output modules for transmission before the transmission.
  • the secure multicomputer system RS2 is to transmit data telegrams to the multicomputer system RS1 via the bus system B1, B2.
  • the RS2 multicomputer system may determine deviations in the data. The output of this data is then prevented and the computer system tries - if necessary after generating the corresponding data again - to determine in a manner yet to be explained which of the data records present in its input / output modules is faulty.
  • the secure multicomputer system then causes a corresponding blocking mark to ensure that no further data are output via this input / output module; if necessary, the recording of data via this input / output module is also blocked.
  • the multicomputer system RS2 teaches the other multicomputer systems from this measure and communicates with these computer systems from now on in the manner explained in more detail below.
  • FIG. 3 shows in FIG. 3 a functional diagram for explaining the method according to the invention for operating a multi-computer system which is formed by two individual computers R1 and R2.
  • the data of interest here is input and output via two signal / input modules E / A1 and E / A2, which are not signal-safe and are routed to separate buses B1 and B2.
  • These input / output modules can be used to receive data telegrams from the computer system and to output data to other computers or to peripheral devices.
  • the data is composed of useful and test data.
  • both computers subject the data supplied to them by both input / output modules to a redundancy check. Both computers should agree that the data record in an input / output module meets the redundancy conditions, but the other does not. In the assumed case, both computers should have recognized the data present in the input / output module E / A2 as faulty, while the data pending in the input / output module E / A1 should meet the specified redundancy conditions.
  • both computers After comparing their test results, both computers cause lock markings to be set and these lock markings prevent the data stored in the input / output module E / A2 from at least during data traffic with the element of the transmission system from which the data was received or stored incorrectly be evaluated or advanced. This happens, for example, in that the computer calls the input / output module recognized as faulty assigned address is prevented.
  • the computer system R1, R2 which detects the malfunction now uses the bus system to inform the data source from which the data causing the malfunction originate about the malfunction that has occurred and the transmission channel B1 affected by the malfunction. The data source in question then in turn blocks the evaluation or switching of data telegrams via the disturbed transmission channel B1.
  • the two computer systems directly affected by the malfunction then only communicate via at least one transmission channel not affected by the malfunction, in the assumed example via bus B2.
  • at least one of the two multi-computer systems directly affected by the fault is to communicate with other computer systems via the bus system, it is advantageous to also inform these computers that certain input / output modules are no longer available for data transmission.
  • the computer system recognizing the malfunction prompts the computer system from which the faulty data originate, or also all computer systems with which it has to communicate, to transmit data to the computer system R1, R2 in the future create more redundant data and only transmit it to the input / output module E / A1; Accordingly, the secure multicomputer system R1, R2 informs the other multicomputer system (s) that it will also create more redundant data itself and will transmit this to the multicomputer systems via the input / output module E / A1.
  • Highly redundant data are to be understood as so-called long telegrams, each consisting of two telegrams of the same length with the same content and being transmitted successively over the remaining channel or channels.
  • the long telegrams to be transmitted after the failure of an input / output module have the same format as the previously transmitted data telegrams, but that they are transmitted twice from the respective data source, in which case their content but their representation does not change.
  • An inverse representation of the data in successive data telegrams is preferably contemplated. Due to the double transmission of data telegrams with the same content, a higher redundancy is achieved compared to a single data transmission.
  • the secure multi-computer system which acts as the receiver, checks whether the content of the user and test information in the data telegrams transmitted one after the other is identical, and whether the transmitted test information matches the test information simulated by the computer system.
  • the evaluation of the data from a computer system with a defective input / output module is only permitted as long as these requirements are met.
  • a redundancy check leads to the result that either the one after the other transmitted user and test information do not match or that the assignment of user and test information is no longer given, the further data traffic to or from the multi-computer system is terminated via the input / output modules which are regarded as faulty.
  • the multi-computer system with the failed input / output modules switches off from the computer network only to the extent that the data transmission via the faulty input / output modules is affected; otherwise, it continues to work and converts messages supplied via other input / output modules into appropriate commands and transfers these commands to the process periphery.
  • the continuation of the data traffic via only one of two input / output modules that are usually supplied with the same data telegrams is, however, not permissible for an unlimited time according to the teaching of the present invention, but this time is limited to a maximum permissible period. If the fault by exchanging or maintaining the faulty input / output module or the transmission channel has not been remedied by this point, further data transmission via the bus system is prohibited.
  • the achievable maximum time in which the telegram transmission may be operated, for example, in a single channel is in a typical computer arrangement with a multitude of signaling-safe multi-computer systems in the order of more than 10 hours. During this time, which can be determined by calculation, the defect that has occurred can be easily remedied by repair or replacement of modules.
  • an input / output module that is safe in terms of signaling is assigned to the safe multi-computer system.
  • This time element which is secure in terms of signal technology and which is illustrated in the drawing by the time elements T1 and T2, is set as soon as the computers of the multi-computer system permanently prevent the evaluation of the data stored in an input / output module. If the switching time impressed on the timer is exceeded, the computers also block the evaluation of the still intact input / output module. This blocking also occurs if at least one of the individual computers detects any errors in a long telegram communicated to it and if these errors cannot be eliminated by repeated data transmission.
  • the multi-computer system concerned For the continuation of the data traffic after the occurrence of a fault in an input / output module, it is crucial that the multi-computer system concerned actually recognizes the defective input / output module or the faulty data channel and does not take over data from the defective input / output module or from the faulty data channel . As already explained, the defective and the intact input / output module are determined via redundancy checks.
  • the methods according to the invention for operating a multi-computer system that is secure in terms of signaling technology are not limited to the design of the secure multi-computer system as a two-by-two / computer system. Rather, they can be used to advantage with any signaling-safe multi-computer system that uses at least two input / output modules with the same data telegrams with other computers or with the periphery communicates. With such a complex multi-computer system, it may be permissible not only to tolerate the failure of a single input / output module, but also, if necessary, the failure of several input / output modules if correct data traffic is still possible via the existing intact input / output modules.
  • the methods according to the invention can also be used advantageously in particular in a two-by-three / computer system.
  • Any transmission systems can be provided for the data transmission between the individual multicomputer systems; however, preference is given to using linear bus systems, in particular ring bus systems.

Abstract

The multi-processor system (RS1) continuously checks the data items transmitted via separate channels and which are present in the input/output modules (E/A11,E/A12) of its individual processors, for correspondence. When deviations occur, the multi-processor system carries out redundancy checks and determines from the result of these checks the input/output module (for example E/A11) which is defective in each case. The data traffic is then handled for a particular time via an input/output module (E/A12) which is still intact. During this time, data items are transmitted with higher redundancy than during the correct operating state of all input/output modules. <IMAGE>

Description

Die Erfindung bezieht sich auf ein Verfahren zum Betrieb eines signaltechnisch sicheren Mehrrechnersystems mit mindestens zwei signaltechnisch nicht sicheren Ein/Ausgabebaugruppen, über die die Rechner des Mehrrechnersystems mindestens zweikanalig Datentelegramme von und/oder zu anderen Rechnern und/oder sonstigen Daten aufnehmenden, abgebenden oder verarbeitenden Systemen, übertragen und über die sie von dort in Form von inhaltlich übereinstimmenden Telegrammen für die Ein/Ausgabebaugruppen durch Prüfdaten gesicherte Nutzdaten empfangen bzw. nach dort abgeben.The invention relates to a method for operating a multi-computer system that is secure in terms of signaling technology, with at least two input / output modules that are not secure in terms of signaling technology, by means of which the computers of the multi-computer system use at least two-channel data telegrams from and / or to other computers and / or other data-receiving, outputting or processing systems , transmitted and via which they receive or deliver useful data secured by test data for the input / output modules from there in the form of telegrams with the same content.

Für besonders sicherheitsrelevante Anwendungsfälle wie z.B. das Eisenbahnsicherungswesen sind signaltechnisch sichere Mehrrechnersysteme entwickelt worden, in denen die zu verarbeitenden Daten in mehreren Kanälen jeweils unabhängig voneinander nach bestimmten Gesetzmäßigkeiten behandelt werden; durch Vergleichsprozeduren werden etwaige Datenabweichungen in den einzelnen Kanälen erkannt und in entsprechende Steuerkommandos umgesetzt, die im Sinne eines Gefährdungsausschlusses auf den zu steuernden Prozeß wirken. Signaltechnisch sichere Mehrrechnersysteme müssen aber nicht nur in der Lage sein, die ihnen zugeführten Daten in der vorgegebenen Weise zu behandeln, sondern sie müssen auch in der Lage sein festzustellen, ob die in ihren signaltechnisch nicht sicheren Ein/Ausgabebaugruppen anstehenden Daten beispielsweise infolge einer Übertragungsstörung oder eines Bauteiledefektes verfälscht sind oder nicht, um dann die Auswertung dieser verfälschten Daten zu unterbinden.For particularly security-relevant applications, such as railway safety, signal-safe multi-computer systems have been developed in which the data to be processed are treated independently of one another in several channels according to certain laws; Comparative procedures identify any data deviations in the individual channels and convert them into corresponding control commands, which act on the process to be controlled in the sense of a risk exclusion. Multi-computer systems that are secure in terms of signaling not only have to be able to treat the data supplied to them in the specified manner, they also have to be able to determine whether they are not in terms of their signaling data present in safe input / output modules, for example as a result of a transmission fault or a component defect, are not falsified, in order to then prevent the evaluation of these falsified data.

Ein bekanntes Verfahren zur Sicherung der Datenübermittlung sieht vor, den eigentlichen Nutzdaten sendeseitig nach bestimmten Gesetzmäßigkeiten gebildete Prüfdaten beizugeben, aus den übermittelten Nutzdaten empfangsseitig nach den gleichen Gesetzmäßigkeiten die zugehörigen Prüfdaten nachzubilden und diese mit den übermittelten Prüfdaten zu vergleichen. Diese Art der Datensicherung ist im allgemeinen um so wirkungsvoller, je aufwendiger die Prüfdaten werden.A known method for securing the data transmission provides for the test data formed on the transmission side to be added to the actual user data, the corresponding test data to be reproduced on the receiving side according to the same laws on the receiving side and to be compared with the transmitted test data. This type of data backup is generally more effective the more complex the test data becomes.

Bei der heutigen Technik sind die Ein/Ausgabebaugruppen elektronischer Datenverarbeitungseinrichtungen üblicherweise mit hochintegrierten Schaltkreisen (FIFOS) besetzt, deren mögliche Ausfallreaktionen nicht umfassend bekannt sind und aus Aufwandsgründen auch nicht durch Analyse ermittelt werden können. Dies hat zur Folge, daß der Nachweis dafür, daß es keine Ausfallwirkung gibt, die trotz Redundanzprüfung unerkannt bleibt, in letzter Konsequenz nicht vollständig geführt werden kann, so daß auch bei sehr umfangreichen Prüfdaten nicht erkennbare Nutzdatenverfälschungen nicht ausgeschlossen werden können. Deshalb erfüllt dieses Lösungskonzept die Forderungen hinsichtlich signaltechnischer Sicherheit als Voraussetzung für die Zulassung einer Telegrammübertragung für besonders sicherheitsrelevante Anwendungen nur unzureichend.In today's technology, the input / output modules of electronic data processing devices are usually equipped with highly integrated circuits (FIFOS), the possible failure reactions of which are not fully known and, for reasons of complexity, cannot be determined by analysis. As a result, the proof that there is no failure effect that remains undetected despite the redundancy check cannot ultimately be provided in full, so that even with very large test data, undetectable data corruption cannot be ruled out. For this reason, this solution concept only insufficiently fulfills the requirements regarding signal safety as a prerequisite for the approval of a telegram transmission for particularly security-relevant applications.

Nicht vorhersagbare Ausfallreaktionen einer Baugruppe, z.B. einer mit hochintegrierten Schaltkreisen bestückten Ein/Ausgabebaugruppe, lassen sich grundsätzlich durch Vervielfachung dieser Baugruppen und Vergleich ihrer Reaktionen miteinander erkennen. Für den vorliegenden Anwendungsfall bedeutet dies, daß die ein-bzw. auszugebenden Informationen jeweils parallel in mindestens zwei gesonderte Ein/Ausgabebaugruppen einzuschreiben sind und daß das Mehrrechnersystem dann die gegenseitige Übereinstimmung der abgespeicherten Daten überprüft. Wird beim Vergleich der in diesen Baugruppen gespeicherten Daten eine Ungleichheit erkannt, so sperrt das Rechnersystem die Verwertung der gespeicherten Informationen. Damit ist die Ungefährlichkeit des ersten Ausfalls einer Ein/Ausgabebaugruppe absolut gegeben.Unpredictable failure reactions of a module, e.g. one equipped with highly integrated circuits An input / output module can basically be identified by multiplying these modules and comparing their reactions with one another. For the present application, this means that the one or. Information to be output in each case must be written in parallel in at least two separate input / output modules and that the multi-computer system then checks the mutual agreement of the stored data. If an inequality is recognized when comparing the data stored in these modules, the computer system blocks the use of the stored information. This means that the first failure of an input / output module is absolutely harmless.

Ein Defekt an einer Ein/Ausgabebaugruppe soll zwar als solcher sofort erkannt und in seiner Auswirkung unwirksam gemacht werden; ein derartiger Defekt soll jedoch nach Möglichkeit nicht den weiteren Datenverkehr des Rechnersystems mit den Systemelementen blockieren, die die durch den Defekt betroffene Ein/Ausgabebaugruppe mit Daten versorgen bzw. von dort Daten beziehen und er soll insbesondere nicht zum Ausfall des gesamten Rechnersystems führen. Um den Datenverkehr auch beim Ausfallen einzelner Ein/Ausgabebaugruppen fortführen zu können, ist es bekannt, jeder Ein/Ausgabebaugruppe mindestens eine entsprechende Reservebaugruppe beizuordnen, die spätestens im Störungsfall zu aktivieren ist und dann die Datenübertragung übernimmt.A defect in an input / output module should be recognized as such immediately and its effects rendered ineffective; However, such a defect should, if possible, not block the further data traffic of the computer system with the system elements which supply the input / output module affected by the defect with data or obtain data from there, and in particular should not lead to failure of the entire computer system. In order to be able to continue the data traffic even if individual input / output modules fail, it is known to assign at least one corresponding reserve module to each input / output module, which is to be activated at the latest in the event of a fault and then takes over the data transmission.

Nachteilig an dieser Ausführung ist der hohe Aufwand für die Ein/Ausgabebaugruppen. Dies ist besonders schwerwiegend, wenn es sich bei den Ein/Ausgabebaugruppen um die Schnittstellen zu vielen anderen Rechnern oder sonstigen Anlagenelementen mit gleichen Anforderungen handelt. Dann muß nämlich dieser zusätzliche Aufwand an allen Rechnern getrieben werden und das geht erheblich ins Geld.A disadvantage of this design is the high effort for the input / output modules. This is particularly serious if the input / output modules are the interfaces to many other computers or other system elements with the same requirements. Then this additional effort has to be driven on all computers and this is very expensive.

Eine Möglichkeit, die beiden vorgenannten Sicherungsmethoden miteinander zu verknüpfen, besteht darin, die zu übermittelnden Daten mit Prüfdaten zu versehen und mehrkanalig auszuwerten. Dieses Verfahren wird bei der in der DE-OS 26 09 107 offenbarten Schaltung angewandt. Nach Prüfung der Redundanz in mehreren voneinander unabhängigen Baugruppen werden die von diesen Baugruppen gebildeten Ergebnisse miteinander verglichen. Wenn man davon ausgeht, daß - wie bereits dargelegt - durch Redundanzprüfungen Datenverfälschungen in einem Register nicht mit absoluter Sicherheit erkannt werden können, dann können derartige Verfälschungen auch nicht durch eine verdoppelte Redundanzprüfung sicher aufgezeigt werden. Das bedeutet, daß die aus der DE-OS 26 09 107 bekannte Schaltung den hohen Sicherheitsanforderungen signaltechnisch sicherer Mehrrechnersysteme nicht gerecht wird.One way of linking the two aforementioned backup methods to one another is to provide the data to be transmitted with test data and to evaluate them in a multi-channel manner. This method is used in the circuit disclosed in DE-OS 26 09 107. After checking the redundancy in several mutually independent modules, the results formed by these modules are compared with one another. If one assumes that - as already explained - data falsifications in a register cannot be detected with absolute certainty by redundancy checks, then such falsifications cannot be reliably identified by a double redundancy check. This means that the circuit known from DE-OS 26 09 107 does not meet the high security requirements of signal-safe multi-computer systems.

Aufgabe der vorliegenden Erfindung ist es, Verfahren zum Betrieb eines signaltechnisch sicheren Mehrrechnersystems gemäß dem Oberbegriff des Patentanspruches 1 bzw. 2 anzugeben, die etwaige Datenverfälschungen in bestimmten, durch inhaltlich gleiche Datentelegramme beschickten Ein/Ausgabebaugruppen sicher erkennen, um sie in ihrer Auswirkung unwirksam zu machen und die es darüber hinaus gestatten , den Datenverkehr auch nach dem Auftreten einer Datenverfälschung mindestens bedingt aufrechtzuerhalten, ohne daß es hierzu der Bereitstellung zusätzlicher Reservebaugruppen bedarf. Die erfindungsgemäßen Verfahren sollen dabei in sich so transparent sein, daß der Nachweis für den sicheren Aufgriff von Datenverfälschungen als Voraussetzung für die Zulassung der erfindungsgemäßen Verfahren bei sicherheitsrelevanten Anwendungen möglich ist.It is an object of the present invention to provide methods for operating a multi-computer system which is secure in terms of signal technology, according to the preamble of patent claim 1 or 2, which reliably detect any data falsifications in certain input / output modules supplied with the same data telegrams in order to render them ineffective and which, moreover, allow data traffic to be maintained at least to a limited extent even after data corruption has occurred, without the need to provide additional reserve modules. The methods of the invention should be so transparent in themselves that it is possible to prove that data corruption has been safely accessed as a prerequisite for the approval of the method according to the invention in security-relevant applications.

Die Erfindung löst diese Aufgabe durch die kennzeichnenden Merkmale des Patentanspruches 1 bzw. die des Patentanspruches 2. Vorteilhafte Aus- und Weiterbildungen der Erfindung sind in den Unteransprüchen angegeben.The invention solves this problem by the characterizing features of claim 1 and that of claim 2. Advantageous developments and further developments of the invention are specified in the dependent claims.

Beide Verfahren sind so beschaffen, daß sie auf Datenverfälschungen in den mit übereinstimmenden Datentelegrammen zu belegenden Ein/Ausgabebaugruppen eines Mehrrechnersystems sicher reagieren, indem sie die Auswertung der fehlerbehafteten Daten verhindern und die an der Datenverfälschung tatsächlich oder möglicherweise beteiligten Elemente von der weiteren Datenübermittlung ausschließen. Die weitere Übermittlung von Daten geschieht in Form sogenannter Langtelegramme, die aus jeweils zwei inhaltlich übereinstimmenden Telegrammen üblicher Länge bestehen, ausschließlich über solche Ein/Ausgabebaugruppen und solche Übertragungskanäle, die durch die festgestellte Störung nicht betroffen sind. Dabei führen die beiden Verfahren zu unterschiedlichen Reaktionen bezogen auf die Anzahl der von der jeweiligen Störung betroffenen Systemelemente: Nach den Merkmalen des Anspruches 1 bleibt die jeweilige Störung mindestens zunächst auf diejenigen Systemelemente beschränkt, die an einer fehlerhaften Datenübertragung tatsächlich beteiligt waren, also auf eine Aus- und eine Eingabebaugruppe zweier Systemelemente; die Übertragung von Daten zwischen allen übrigen Systemelementen, auch zwischen den durch die Störung betroffenen und den übrigen Systemelementen, geschieht vorerst - bis zum Auftreten eventueller weiterer Störungen - zweikanalig durch Telegramme üblicher Länge, danach bei dem durch diese weiteren Störungen betroffenen Systemelementen ebenfalls einkanalig mittels Langtelegrammen. Demgegenüber sehen die Merkmale des Anspruches 2 vor, beim Auftreten einer Störung alle Systemelemente hiervon zu unterrichten und dafür zu sorgen, daß alle Systemelemente im Verkehr mit den durch die Störung direkt betroffenen Systemelementen fortan z.B. einkanalig unter Verwendung von Langtelegrammen verkehren.Both methods are designed in such a way that they reliably react to data corruption in the input / output modules of a multicomputer system to be filled with matching data telegrams by preventing the evaluation of the faulty data and excluding the elements that are or possibly involved in the data corruption from further data transmission. The further transmission of data takes place in the form of so-called long telegrams, each consisting of two telegrams of the same length, each with the same content, exclusively via those input / output modules and transmission channels that are not affected by the detected fault. The two methods lead to different reactions in relation to the number of system elements affected by the respective disturbance: According to the features of claim 1, the respective disturbance remains at least initially restricted to those system elements that were actually involved in a faulty data transmission, that is, to an off - And an input module of two system elements; The transmission of data between all other system elements, including those affected by the malfunction and the other system elements, takes place initially - until any further malfunctions occur - using two-channel telegrams of normal length, then also in single-channel long telegrams for the system elements affected by these malfunctions . In contrast, the features of claim 2 provide to inform all system elements when a fault occurs and to ensure that all system elements in traffic with the system elements directly affected by the fault henceforth, for example, use single-channel using long telegrams.

Nach beiden Verfahren wird die Datenübertragung im Anschluß an das Auffinden fehlerbehafteter Daten unter zusätzlicher Datensicherung und Belegung eines durch die Datenverfälschung nicht betroffenen Übertragungskanals für eine bestimmte Zeitspanne zugelassen, die zum Beheben des eingetretenen Defektes genutzt werden kann.According to both methods, the data transmission is permitted for a certain period of time after the finding of faulty data with additional data backup and allocation of a transmission channel not affected by the data corruption, which can be used to remedy the defect that has occurred.

Die Erfindung ist nachstehend anhand von in der Zeichnung schematisch dargestellten Ausführungsbeispielen näher erläutert.The invention is based on exemplary embodiments shown schematically in the drawing explained in more detail.

Die Zeichnung zeigt in den Figuren 1 und 2 einen besonders vorteilhaften Aufbau von aus mehreren signaltechnisch sicheren Mehrrechnersystemen bestehenden Rechneranordnungen, die in unterschiedlicher Weise auf fehlerhaft abgelegte Daten reagieren, und in Figur 3 ein Funktionsschaubild zur Erläuterung der erfindungsgemäßen Verfahren bei einem signaltechnisch sicheren Zweirechnersystem.The drawing shows in FIGS. 1 and 2 a particularly advantageous structure of computer arrangements consisting of several multi-computer systems which are secure in terms of signaling and which react in different ways to incorrectly stored data, and in FIG. 3 is a functional diagram for explaining the methods according to the invention in a two-computer system which is secure in terms of signaling.

Die in Figur 1 dargestellte Rechneranordnung besteht aus beispielsweise sechs signaltechnisch sicheren Mehrrechnersystemen RS1 bis RS6, die über zugeordnete Ein/Ausgabebaugruppen an die beiden Busse B1 und B2 eines Bussystems angeschlossen sind, über das die Mehrrechnersysteme zweikanalig miteinander kommunizieren. Es ist angenommen, daß das Rechnersystem RS2 Daten in Form eines Telegrammes an das Rechnersystem RS1 zu übermitteln hat. Hierzu aktiviert das Rechnersystem RS2 seine beiden Ein/Ausgabebaugruppen E/A21 und E/A22 zur beispielsweise adressenorientierten Übertragung inhaltlich gleicher Datentelegramme über die beiden Busse B1 und B2 an die Ein/Ausgabebaugruppen E/A11 und E/A12 des Rechnersystems RS1; die über die beiden Busse übertragenen Datentelegramme können dabei unterschiedlich, z.B. invers dargestellt sein. Das die Daten aufnehmende Rechnersystem RS1 prüft die in seinen Ein/Ausgabebaugruppen abgelegten Daten vor ihrer Anerkennung auf Einhaltung des festgelegten Datenformates und auf inhaltliche Übereinstimmung. Dabei möge das Rechnersystem RS1 in seinen Ein/Ausgabebaugruppen E/A11 und E/A12 voneinander abweichende Daten feststellen. Das Rechnersystem verwirft daraufhin das übermittelte Datentelegramm und fordert über beide Busse B1, B2 vom Datensender, dem sicheren Mehrrechnersystem RS2, eine erneute Telegrammübertragung an. Genügen die dabei übermittelten Daten den Prüfbedingungen des Rechnersystems RS1, so werden sie anerkannt, wobei die Anerkennung der Daten dem Datensender quittiert werden kann. Endet auch der zweite Versuch einer Datenübermittlung negativ, d.h. stellt der Datenempfänger inhaltlich oder vom Format her abweichende Daten in seinen Ein/Ausgabebaugruppen fest, so stellt er auf noch zu erläuternde Weise fest, welcher der in seinen Ein/Ausgabebaugruppen abgelegten Datensätze fehlerhaft ist und welcher nicht. Im vorliegenden Beispiel möge das Rechnersystem RS1 die in seiner Ein/Ausgabebaugruppe E/A11 anstehenden Daten als fehlerhaft erkannt haben. Es sperrt daraufhin rechnerintern die weitere Aufnahme von Daten über die betreffende Ein/Ausgabebaugruppe und unterrichtet mindestens das Mehrrechnersystem RS2, von dem das fehlerbehaftete Datentelegramm stammte, hiervon. Dies führt dort zum Setzen eines entsprechenden Sperrvermerkes, der die weitere Übertragung von Daten über die Ein/Ausgabebaugruppe E/A21 zur Ein/Ausgabebaugruppe E/A11 des Mehrrechnersystems RS1 verhindert. Damit wird dem Umstand Rechnung getragen, daß empfangsseitig zwar die aufgetretene Datenverfälschung erkennbar ist, nicht jedoch, wo der Ort dieser Datenverfälschung liegt.The computer arrangement shown in FIG. 1 consists, for example, of six multi-computer systems RS1 to RS6 which are secure in terms of signal technology and which are connected to the two buses B1 and B2 of a bus system via assigned input / output modules, via which the multi-computer systems communicate with one another in two channels. It is assumed that the computer system RS2 has to transmit data to the computer system RS1 in the form of a telegram. For this purpose, the computer system RS2 activates its two input / output modules E / A21 and E / A22 for, for example, address-oriented transmission of data telegrams of the same content via the two buses B1 and B2 to the input / output modules E / A11 and E / A12 of the computer system RS1; The data telegrams transmitted via the two buses can be displayed differently, for example, inversely. The RS1 computer system that records the data checks the data stored in its input / output modules before they are recognized for compliance with the specified data format and for content-related correspondence. The RS1 computer system may determine different data in its input / output modules E / A11 and E / A12. The computer system then discards the transmitted data telegram and requests both Buses B1, B2 from the data transmitter, the secure multi-computer system RS2, send a new telegram transmission. If the data transmitted thereby satisfy the test conditions of the RS1 computer system, they are recognized, and the data sender can be acknowledged if the data are recognized. If the second attempt at data transmission also ends negatively, i.e. if the data recipient detects data in its input / output modules that differ in content or format, it will determine in a manner yet to be explained which of the data records stored in its input / output modules is faulty and which one Not. In the present example, the computer system RS1 may have recognized the data pending in its input / output module I / O11 as incorrect. It then blocks the further recording of data via the input / output module in question and informs at least the multicomputer system RS2 from which the faulty data telegram originated. This leads to the setting of a corresponding blocking note, which prevents the further transmission of data via the input / output module E / A21 to the input / output module E / A11 of the multicomputer system RS1. This takes into account the fact that the data corruption that has occurred can be recognized at the receiving end, but not where the location of this data corruption is.

Wenn die von der festgestellten Störung direkt betroffenen beiden Rechnersysteme auch mit anderen, an das gleiche Bussystem angeschlossenen Rechnersystemen kommunizieren sollen, ist es vorteilhaft, auch diese Rechner davon zu unterrichten, daß bestimmte Ein/Ausgabebaugruppen des die Störung feststellenden Rechnersystems und ggf. des Rechnersystems, von dem das verfälschte Datentelegramm stammte, für die weitere Datenübertragung nicht mehr zur Verfügung stehen. Hierdurch soll verhindert werden, daß ein Defekt in einer einzigen Ein/Ausgabebaugruppe nach und nach zum Sperren weiterer Ein/Ausgabebaugruppen anderer Rechnersysteme führt.If the two computer systems directly affected by the detected fault are also to communicate with other computer systems connected to the same bus system, it is advantageous to also inform these computers that certain input / output modules of the computer system which is detecting the fault and possibly the computer system, from which the corrupted data telegram originated for further data transmission are no longer available. This is to prevent a defect in a single input / output module from gradually blocking further input / output modules from other computer systems.

Die durch die Störung direkt betroffenen Mehrrechnersysteme RS1 und RS2 korrespondieren untereinander und ggf. mit den durch die Störung nicht direkt betroffenen Mehrrechnersystemen RS3 bis RS6 fortan auf eine Weise, wie sie im einzelnen anhand der Figur 3 erläutert wird.The multicomputer systems RS1 and RS2 which are directly affected by the disturbance correspond to one another and possibly with the multicomputer systems RS3 to RS6 which are not directly affected by the disturbance in a manner as will be explained in detail with reference to FIG. 3.

Bei dem Ausführungsbeispiel der Figur 2 ist angenommen, daß auch das jeweils sendende Mehrrechnersystem die in seinen Ein/Ausgabebaugruppen zur Übertragung anstehenden Daten vor der Übertragung auf Übereinstimmung prüft. Wieder soll dabei das sichere Mehrrechnersystem RS2 Datentelegramme über das Bussystem B1, B2 an das Mehrrechnersystem RS1 übermitteln. Beim Rücklesen der in den Ein/Ausgabebaugruppen E/A21 und E/A22 anstehenden Daten möge das Mehrrechnersystem RS2 Abweichungen in den Daten feststellen. Die Ausgabe dieser Daten wird daraufhin verhindert und das Rechnersystem versucht - ggf. nach erneuter Generierung der entsprechenden Daten - auf noch zu erläuternde Weise festzustellen, welcher der in seinen Ein/Ausgabebaugruppen anliegenden Datensätze fehlerhaft ist. Es möge dabei zu der Erkenntnis gelangen, daß die in seiner Ein/Ausgabebaugruppe E/A21 anliegenden Daten fehlerhaft sind. Das sichere Mehrrechnersystem veranlaßt daraufhin durch Setzen einer entsprechenden Sperrmarkierung, daß über diese Ein/Ausgabebaugruppe keine weiteren Daten abgegeben werden; ggf. wird auch die Aufnahme von Daten über diese Ein/Ausgabebaugruppe gesperrt. Das Mehrrechnersystem RS2 unterrichtet die übrigen Mehrrechnersysteme von dieser Maßnahme und kommuniziert mit diesen Rechnersystemen fortan auf die nachstehend näher erläuterte Weise.In the exemplary embodiment in FIG. 2, it is assumed that the transmitting multicomputer system in each case also checks the data pending in its input / output modules for transmission before the transmission. Again, the secure multicomputer system RS2 is to transmit data telegrams to the multicomputer system RS1 via the bus system B1, B2. When reading back the data pending in the input / output modules E / A21 and E / A22, the RS2 multicomputer system may determine deviations in the data. The output of this data is then prevented and the computer system tries - if necessary after generating the corresponding data again - to determine in a manner yet to be explained which of the data records present in its input / output modules is faulty. It may come to the conclusion that the data present in its input / output module E / A21 are incorrect. The secure multicomputer system then causes a corresponding blocking mark to ensure that no further data are output via this input / output module; if necessary, the recording of data via this input / output module is also blocked. The multicomputer system RS2 teaches the other multicomputer systems from this measure and communicates with these computer systems from now on in the manner explained in more detail below.

Die Zeichnung zeigt in Figur 3 ein Funktionsschaubild für die Erläuterung der erfindungsgemäßen Verfahren zum Betrieb eines Mehrrechnersystems, das durch zwei Einzelrechner R1 und R2 gebildet wird. In der Zeichnung sind nur diejenigen Elemente dargestellt, die für die Erläuterung der erfindungsgemäßen Verfahren erforderlich sind. Die Ein- bzw. Ausgabe der hier interessierenden Daten erfolgt über zwei signaltechnisch nicht sichere Ein/Ausgabebaugruppen E/A1 und E/A2, die auf gesonderte Busse B1 und B2 geführt sind. Über diese Ein/Ausgabebaugruppen können sowohl Datentelegramme vom Rechnersystem aufgenommen als auch Daten an andere Rechner oder an eine Peripherie abgegeben werden. Die Daten setzen sich aus Nutz- und Prüfdaten zusammen.The drawing shows in FIG. 3 a functional diagram for explaining the method according to the invention for operating a multi-computer system which is formed by two individual computers R1 and R2. In the drawing only those elements are shown which are necessary for the explanation of the method according to the invention. The data of interest here is input and output via two signal / input modules E / A1 and E / A2, which are not signal-safe and are routed to separate buses B1 and B2. These input / output modules can be used to receive data telegrams from the computer system and to output data to other computers or to peripheral devices. The data is composed of useful and test data.

Die bei der Eingabe von Daten in das Zweirechnersystem über die Busse B1, B2 zweikanalig angelieferten und in den Registern der Ein/Ausgabebaugruppen E/A1 und E/A2 anstehenden Daten werden im Zweirechnersystem jeweils vor ihrer Anerkennung miteinander verglichen, wobei sich dieser Vergleich zunächst nur auf die Nutzdaten beschränken kann; es ist aber auch möglich, die jeweils übermittelten Daten grundsätzlich auch einer Redundanzprüfung zu unterziehen, d.h. die Zuordnung von Nutz- und Prüfdaten zu prüfen. Dieser Vergleich bzw. die zusätzliche Redundanzprüfung findet dezentral in den beiden Einzelrechnern R1, R2 des sicheren Mehrrechnersystem statt. Jeder Einzelrechner vergleicht bzw. prüft die ihm von beiden Ein/Ausgabebaugruppen E/A1 und E/A2 angebotenen Datensätze. Stellen die Rechner unabhängig voneinander aber übereinstimmend abweichende Daten in den beiden Ein/Ausgabebaugruppen fest, so verwerfen sie die dort abgelegten Daten und fordern vom jeweiligen Datensender die erneute Übertragung der von ihnen als fehlerhaft erkannten Daten an. Genügen diese Daten den gegebenen Bewertungsbedingungen, so werden sie von den beiden Einzelrechnern akzeptiert und der weiteren Verarbeitung RV1, RV2 zugeführt.The data supplied when entering data in the two-computer system via buses B1, B2 in two channels and pending in the registers of the input / output modules E / A1 and E / A2 are compared in the two-computer system before they are recognized, whereby this comparison is initially only can limit to the user data; However, it is also possible to fundamentally subject the data transmitted in each case to a redundancy check, ie to check the allocation of useful and test data. This comparison or the additional redundancy check takes place decentrally in the two individual computers R1, R2 of the secure multi-computer system. Each individual computer compares or checks the data records offered by both input / output modules E / A1 and E / A2. Place the computers independently of each other However, if the data in the two input / output modules are the same, they discard the data stored there and request the respective data sender to retransmit the data they have identified as incorrect. If these data meet the given evaluation conditions, they are accepted by the two individual computers and passed on for further processing RV1, RV2.

Genügen die angeforderten Daten auch jetzt den gegebenen Bewertungsbedingungen nicht, so werden auch sie verworfen. Die beiden Einzelrechner R1, R2 versuchen nun festzustellen, welche der beiden Ein/Ausgabebaugruppen fehlerhafte Daten enthält und welche nicht. Hierzu unterziehen beide Rechner die ihnen von beiden Ein/Ausgabebaugruppen zugeführten Daten einer Redundanzprüfung. Dabei sollen beide Rechner übereinstimmend feststellen, daß der in einer Ein/Ausgabebaugruppe anstehende Datensatz den Redundanzbedingungen genügt, der andere jedoch nicht. Im angenommenen Fall sollen beide Rechner die in der Ein/Ausgabebaugruppe E/A2 anliegenden Daten als fehlerhaft erkannt haben, während die in der Ein/Ausgabebaugruppe E/A1 anstehenden Daten den vorgegebenen Redundanzbedingungen genügen sollen.If the requested data still does not meet the given evaluation conditions, it will also be discarded. The two individual computers R1, R2 are now trying to determine which of the two input / output modules contains incorrect data and which does not. For this purpose, both computers subject the data supplied to them by both input / output modules to a redundancy check. Both computers should agree that the data record in an input / output module meets the redundancy conditions, but the other does not. In the assumed case, both computers should have recognized the data present in the input / output module E / A2 as faulty, while the data pending in the input / output module E / A1 should meet the specified redundancy conditions.

Nach dem Vergleich ihrer Prüfergebnisse veranlassen beide Rechner das Setzen von Sperrmarkierungen und diese Sperrmarkierungen verhindern, daß die in der Ein/Ausgabebaugruppe E/A2 abgelegten Daten mindestens beim Datenverkehr mit demjenigen Element des Übertragungssystems, von dem die Daten fehlerhaft empfangen bzw. abgelegt wurden, fortan ausgewertet bzw. fortgeschaltet werden. Dies geschieht z.B. dadurch, daß rechnerintern der Aufruf der der als gestört erkannten Ein/Ausgabebaugruppe zugeordneten Adresse verhindert wird.
Das die Störung erkennende Rechnersystem R1, R2 unterrichtet nun über das Bussystem die Datenquelle, von der die die Störung verursachenden Daten stammen, über die eingetretene Störung und den durch die Störung betroffenen Übertragungskanal B1. Die jeweils angesprochene Datenquelle sperrt daraufhin ihrerseits die Auswertung bzw. Fortschaltung von Datentelegrammen über den gestörten Übertragungskanal B1. Die beiden durch die Störung direkt betroffenen Rechnersysteme kommunizieren dann nur noch über mindestens einen durch die Störung nicht betroffenen Übertragungskanal, im angenommenen Beispiel über den Bus B2.
Für den Fall, daß mindestens einer der beiden von der Störung direkt betroffenen Mehrrechnersysteme über das Bussystem noch mit weiteren Rechnersystemen kommunizieren soll, ist es vorteilhaft, auch diese Rechner davon zu unterrichten, daß bestimmte Ein/Ausgabebaugruppen für die Datenübertragung nicht mehr zur Verfügung stehen. Gleichzeitig mit dem Setzen der Sperrmarkierung fordert das die Störung erkennende Rechnersystem das Rechnersystem, von dem die fehlerbehafteten Daten stammen, bzw. auch alle Rechnersysteme, mit denen es zu kommunizieren hat, dazu auf, zukünftig für die Übertragung von Daten an das Rechnersystem R1, R2 höherredundante Daten zu erstellen und ausschließlich an die Ein/Ausgabebaugruppe E/A1 zu übertragen; entsprechend unterrichtet das sichere Mehrrechnersystem R1, R2 das bzw. die übrigen Mehrrechnersysteme davon, daß es selbst ebenfalls höherredundante Daten erstellen und diese über die Ein/Ausgabebaugruppe E/A1 an die Mehrrechnersysteme übermitteln wird. Unter höherredundanten Daten sind sogenannte Langtelegramme zu verstehen, die aus je zwei inhaltlich gleichen Telegrammen üblicher Lange bestehen und nacheinander über den oder die verbliebenen Kanäle übertragen werden.After comparing their test results, both computers cause lock markings to be set and these lock markings prevent the data stored in the input / output module E / A2 from at least during data traffic with the element of the transmission system from which the data was received or stored incorrectly be evaluated or advanced. This happens, for example, in that the computer calls the input / output module recognized as faulty assigned address is prevented.
The computer system R1, R2 which detects the malfunction now uses the bus system to inform the data source from which the data causing the malfunction originate about the malfunction that has occurred and the transmission channel B1 affected by the malfunction. The data source in question then in turn blocks the evaluation or switching of data telegrams via the disturbed transmission channel B1. The two computer systems directly affected by the malfunction then only communicate via at least one transmission channel not affected by the malfunction, in the assumed example via bus B2.
In the event that at least one of the two multi-computer systems directly affected by the fault is to communicate with other computer systems via the bus system, it is advantageous to also inform these computers that certain input / output modules are no longer available for data transmission. Simultaneously with the setting of the blocking marking, the computer system recognizing the malfunction prompts the computer system from which the faulty data originate, or also all computer systems with which it has to communicate, to transmit data to the computer system R1, R2 in the future create more redundant data and only transmit it to the input / output module E / A1; Accordingly, the secure multicomputer system R1, R2 informs the other multicomputer system (s) that it will also create more redundant data itself and will transmit this to the multicomputer systems via the input / output module E / A1. Highly redundant data are to be understood as so-called long telegrams, each consisting of two telegrams of the same length with the same content and being transmitted successively over the remaining channel or channels.

Die vorstehend erläuterten Maßnahmen können sinngemäß auch für einen Datensender gelten, von dem die als fehlerhaft erkannten Daten stammen.The measures explained above can also apply mutatis mutandis to a data transmitter from which the data identified as incorrect originate.

Einzelne oder alle Mehrrechnersysteme überprüfen nun fortlaufend bei Datenverkehr mit dem bzw. den durch die Störung direkt betroffenen Mehrrechnersystemen den Empfang von Langtelegrammen und prüfen diese auf des Einhalten der Redundanzbedingungen und sie erstellen auch selbst für den Datenverkehr mit diesen Mehrrechnersystemen Langtelegramme. Entsprechendes gilt für die Mehrrechnersysteme mit den ausgefallenen Ein/Ausgabeeinrichtungen.Individual or all multicomputer systems now continuously check the receipt of long telegrams in the case of data traffic with the multicomputer system (s) directly affected by the fault and check that the redundancy conditions are met and they also create long telegrams for data traffic with these multicomputer systems. The same applies to the multi-computer systems with the failed input / output devices.

Nach der Erfindung ist vorgesehen daß die nach dem Ausfallen einer Ein/Ausgabebaugruppe zu übertragenden Langtelegramme das gleich Format wie die zuvor übertragenen Datentelegramme aufweisen, daß sie aber von der jeweiligen Datenquelle her zweifach übertragen werden, wobei sich dann nicht ihr Inhalt aber ihre Darstellung ändert. Vorzugsweise ist dabei an eine inverse Darstellung der Daten in aufeinanderfolgenden Datentelegrammen gedacht. Durch die zweimalige Übertragung inhaltlich übereinstimmender Datentelegramme wird gegenüber einer nur einmaligen Datenübertragung eine höhere Redundant erreicht. Das jeweils als Empfänger fungierende sichere Mehrrechnersystem prüft vor der Anerkennung der Daten, ob in den ihm nacheinander übermittelten Datentelegrammen sowohl die Nutz- als auch die Prüfinformationen jeweils von ihrem Inhalt her übereinstimmen und ob die übermittelten Prüfinformationen mit den vom Rechnersystem nachgebildeten Prüfinformationen übereinstimmen. Nur solange diese Voraussetzungen gegeben sind, ist die Auswertung der von einem Rechnersystem mit einer defekten Ein/Ausgabebaugruppe stammenden Daten zulässig. Führt eine Redundanzprüfung zu dem Ergebnis, daß entweder die nacheinander übermittelten Nutz- und Prüfinformationen nicht übereinstimmen oder daß die Zuordnung von Nutz- und Prüfinformationen nicht mehr gegeben ist, so wird der weitere Datenverkehr zum bzw. vom Mehrrechnersystem über die als gestört angesehenen Ein/Ausgabebaugruppen abgebrochen. Das Mehrrechnersystem mit den ausgefallenen Ein/Ausgabebaugruppen schaltet sich aber nur insoweit aus dem Rechnerverbund aus als die Datenübermittlung über die gestörten Ein/Ausgabebaugruppen betroffen ist; im übrigen arbeitet es nach wie vor weiter und setzt über andere Ein/Ausgabebaugruppen zugeführte Meldungen bestimmungsgemäß in entsprechende Kommandos um und überträgt diese Kommandos an die Prozeßperipherie.According to the invention, it is provided that the long telegrams to be transmitted after the failure of an input / output module have the same format as the previously transmitted data telegrams, but that they are transmitted twice from the respective data source, in which case their content but their representation does not change. An inverse representation of the data in successive data telegrams is preferably contemplated. Due to the double transmission of data telegrams with the same content, a higher redundancy is achieved compared to a single data transmission. Before the data is recognized, the secure multi-computer system, which acts as the receiver, checks whether the content of the user and test information in the data telegrams transmitted one after the other is identical, and whether the transmitted test information matches the test information simulated by the computer system. The evaluation of the data from a computer system with a defective input / output module is only permitted as long as these requirements are met. A redundancy check leads to the result that either the one after the other transmitted user and test information do not match or that the assignment of user and test information is no longer given, the further data traffic to or from the multi-computer system is terminated via the input / output modules which are regarded as faulty. The multi-computer system with the failed input / output modules switches off from the computer network only to the extent that the data transmission via the faulty input / output modules is affected; otherwise, it continues to work and converts messages supplied via other input / output modules into appropriate commands and transfers these commands to the process periphery.

Die Fortführung des Datenverkehrs über nur eine von zwei üblicherweise mit inhaltlich gleichen Datentelegrammen versorgte Ein/Ausgabebaugruppen ist nach der Lehre der vorliegenden Erfindung aber nicht für unbegrenzte Zeit zulässig, sondern diese Zeit ist auf eine maximal zulässige Zeitspanne begrenzt. Ist bis zu diesem Zeitpunkt die Störung durch Auswechseln oder Instantsetzen der gestörten Ein/Ausgabebaugruppe bzw. des Übertragungskanals nicht behoben, so verbietet sich die weitere Datenübertragung über das Bussystem. Die maximal zulässige Zeit für das Aufschalten einer einzigen Ein/Ausgabebaugruppe auf beide Einzelrechner des Mehrrechnersystems ist abhängig von der für die betreffende Baugruppe aufgrund ihres Aufbaus zu erwartenden mittleren Zeit zwischen zwei unabhängig voneinander auftretenden Fehlern (MTBF = meantime between failures) und einer Kenngröße, die abhängig ist von der für die Datendarstellung erreichten Redundanz. Die erreichbare maximale Zeit, in der die Telegrammübertragung z.B. einkanalig betrieben werden darf, liegt bei einer typischen Rechneranordnung mit einer Vielzahl von signaltechnisch sicheren Mehrrechnersystemen in der Größenordnung von weit über 10 Stunden. In dieser Zeit, die rechnerisch bestimmt werden kann, läßt sich der eingetretene Defekt durch Instandsetzung oder durch Ersatz von Baugruppen mühelos beheben.The continuation of the data traffic via only one of two input / output modules that are usually supplied with the same data telegrams is, however, not permissible for an unlimited time according to the teaching of the present invention, but this time is limited to a maximum permissible period. If the fault by exchanging or maintaining the faulty input / output module or the transmission channel has not been remedied by this point, further data transmission via the bus system is prohibited. The maximum permissible time for connecting a single input / output module to both individual computers of the multi-computer system depends on the average time to be expected for the module in question due to its structure, between two independently occurring errors (MTBF = meantime between failures) and a parameter that depends on the redundancy achieved for data display. The achievable maximum time in which the telegram transmission may be operated, for example, in a single channel is in a typical computer arrangement with a multitude of signaling-safe multi-computer systems in the order of more than 10 hours. During this time, which can be determined by calculation, the defect that has occurred can be easily remedied by repair or replacement of modules.

Für die Begrenzung der Zeitspanne, für die nach dem Ausfallen einer Ein/Ausgabebaugruppe der Datenverkehr über eine noch intakte Ein/Ausgabebaugruppe weitergeführt werden darf, ist dem sicheren Mehrrechnersystem ein signaltechnisch sicheres Zeitglied zugeordnet. Dieses signaltechnisch sichere Zeitglied, das in der Zeichnung durch die Zeitglieder T1 und T2 verdeutlicht ist, wird eingestellt, sobald die Rechner des Mehrrechnersystems die Auswertung der in einer Ein/Ausgabebaugruppe abgespeicherten Daten bleibend verhindern. Wird die dem Zeitglied eingeprägte Schaltzeit überschritten, so sperren die Rechner auch die Auswertung der noch intakten Ein/Ausgabebaugruppe. Diese Sperrung tritt auch ein, wenn mindestens einer der Einzelrechner in einem ihm mitgeteilten Langtelegramm irgendwelche Fehler entdeckt und wenn diese Fehler nicht durch nochmalige Datenübertragung zu beseitigen sind.In order to limit the period of time for which data traffic may continue after an input / output module has failed, an input / output module that is safe in terms of signaling is assigned to the safe multi-computer system. This time element, which is secure in terms of signal technology and which is illustrated in the drawing by the time elements T1 and T2, is set as soon as the computers of the multi-computer system permanently prevent the evaluation of the data stored in an input / output module. If the switching time impressed on the timer is exceeded, the computers also block the evaluation of the still intact input / output module. This blocking also occurs if at least one of the individual computers detects any errors in a long telegram communicated to it and if these errors cannot be eliminated by repeated data transmission.

Für die Fortführung des Datenverkehrs nach dem Auftreten einer Störung in einer Ein/Ausgabebaugruppe ist ausschlaggebend, daß das betroffene Mehrrechnersystem die defekte Ein/Ausgabebaugruppe beziehungsweise den gestörten Datenkanal auch tatsächlich erkennt und nicht etwa Daten aus der defekten Ein/Ausgabebaugruppe beziehungsweise aus dem gestörten Datenkanal übernimmt. Wie bereits erläutert wird die defekte und die intakte Ein/Ausgabebaugruppe über Redundanzprüfungen ermittelt.For the continuation of the data traffic after the occurrence of a fault in an input / output module, it is crucial that the multi-computer system concerned actually recognizes the defective input / output module or the faulty data channel and does not take over data from the defective input / output module or from the faulty data channel . As already explained, the defective and the intact input / output module are determined via redundancy checks.

Dabei kann der Fall eintreten, daß nicht nur eine, sondern beide Redundanzprüfungen zu dem Ergebnis kommen, daß die jeweils übermittelten Prüfdaten mit den vom Mehrrechnersystem nachgebildeten Prüfdaten übereinstimmen. Dies ist möglich, wenn neben der Verfälschung von Nutzdaten im gleichen Kanal auch die übermittelten Prüfdaten gerade so verfälscht werden, daß sich daraus die für die verfälschten Nutzdaten richtigen Prüfdaten ergeben. Dieser Fall ist zwar sehr unwahrscheinlich, er ist jedoch nicht gänzlich auszuschließen. Tritt dieser Fall ein, so wird dies beim Vergleich der Redundanzprüfergebnisse festgestellt. Da die beiden Einzelrechner aus den Redundanzprüfungen in diesem Fall nicht den gestörten Übertragungskanal eindeutig bestimmen können, unterbinden sie in diesem Fall die weitere Datenübertragung über beide Ein/Ausgabebaugruppen.This can lead to the result that not just one but both redundancy checks come to the result that the test data transmitted in each case match the test data simulated by the multi-computer system. This is possible if, in addition to the falsification of user data in the same channel, the transmitted test data are also falsified in such a way that the correct test data for the falsified user data result. This case is very unlikely, but it cannot be ruled out entirely. If this happens, this is determined by comparing the redundancy test results. Since the two individual computers from the redundancy checks cannot clearly determine the faulty transmission channel in this case, they prevent further data transmission via both input / output modules in this case.

Die vorstehend für Rechneranordnungen nach Figur 1 beschriebenen Abläufe gelten entsprechend auch für Rechneranordnungen gemäß Figur 2, bei denen die Rechner eines Mehrrechnersystems die in ihren Ein/Ausgabebaugruppen zur Übertragung anstehenden Datentelegramme vor ihrer Freigabe zurücklesen und die Freigabe vom Bilden übereinstimmender Prüfergebnisse in den Rechnern des Mehrrechnersystems abhängig machen.The processes described above for computer arrangements according to FIG. 1 also apply accordingly to computer arrangements according to FIG. 2, in which the computers of a multi-computer system read back the data telegrams pending for transmission in their input / output modules before they are released and the release of the formation of matching test results in the computers of the multi-computer system make them dependent.

Die erfindungsgemäßen Verfahren zum Betrieb eines signaltechnisch sicheren Mehrrechnersystems sind nicht auf die Ausbildung des sicheren Mehrrechnersystems als Zweivon-Zwei/Rechnersystem beschränkt. Sie lassen sich vielmehr mit Vorteil bei jedem beliebigen signaltechnisch sicheren Mehrrechnersystem anwenden, das über mindestens zwei mit inhaltlich gleichen Datentelegrammen beschickte Ein/Ausgabebaugruppen mit anderen Rechnern oder mit der Peripherie kommuniziert. Bei einem derartigen aufwendiger gestalteten Mehrrechnersystem kann es zulässig sein, nicht nur den Ausfall einer einzigen Ein/Ausgabebaugruppe, sondern ggf. auch den Ausfall mehrerer Ein/Ausgabebaugruppen zu tolerieren, wenn über die vorhandenen noch intakten Ein/Ausgabebaugruppen ein ordnungsgerechter Datenverkehr noch möglich ist. Die erfindungsgemäßen Verfahren sind insbesondere auch bei einem Zweivon-Drei/Rechnersystem vorteilhaft anwendbar.The methods according to the invention for operating a multi-computer system that is secure in terms of signaling technology are not limited to the design of the secure multi-computer system as a two-by-two / computer system. Rather, they can be used to advantage with any signaling-safe multi-computer system that uses at least two input / output modules with the same data telegrams with other computers or with the periphery communicates. With such a complex multi-computer system, it may be permissible not only to tolerate the failure of a single input / output module, but also, if necessary, the failure of several input / output modules if correct data traffic is still possible via the existing intact input / output modules. The methods according to the invention can also be used advantageously in particular in a two-by-three / computer system.

Für die Datenübertragung zwischen den einzelnen Mehrrechnersystemen können beliebige Übertragungssysteme vorgesehen sein; vorzugsweise ist jedoch an die Verwendung linearer Bussysteme, insbesondere Ringbussysteme gedacht.Any transmission systems can be provided for the data transmission between the individual multicomputer systems; however, preference is given to using linear bus systems, in particular ring bus systems.

Dort, wo derartige Übertragungssysteme verdoppelt werden, geschieht dies überwiegend aus Verfügbarkeitsgründen. Bei dem erfindungsgemäßen Verfahren zum Betrieb eines signaltechnisch sicheren Mehrrechnersystems dagegen wird die Verdoppelung der Bussysteme zunächst ausschließlich zur Erhöhung der Sicherheit beim Datenverkehr zwischen den Rechnersystemen verwendet, indem die Daten mehrkanalig übertragen und die in den Ein/Ausgabebaugruppen jeweils anliegenden Daten vor ihrer Anerkennung auf Übereinstimmung geprüft werden. Erst beim Ausfall eines Übertragungskanals zwischen zwei Mehrrechnersystemen wird die einem verdoppelten Übertragungssystem innenwohnende hohe Verfügbarkeit ausgenutzt und die Datenübertragung vorübergehend einkanalig über das jeweils noch als ordnungsgerecht angesehene Übertragungssystem betrieben; dies ist nach der Lehre der Erfindung zulässig, sofern die dabei übermittelten Daten durch zusätzliche Maßnahmen gegen nicht erkennbare Fehler zusätzlich gesichert werden.Wherever such transmission systems are doubled, this is done primarily for reasons of availability. In the case of the method according to the invention for operating a multi-computer system that is secure in terms of signaling, on the other hand, the doubling of the bus systems is initially used exclusively to increase the security of data traffic between the computer systems by transmitting the data in multiple channels and checking the data present in the input / output modules for compliance before they are recognized will. Only when a transmission channel between two multicomputer systems fails does the high availability inherent in a doubled transmission system be exploited and the data transmission temporarily operated via one channel via the transmission system, which is still considered to be correct; this is permissible according to the teaching of the invention, provided that the data transmitted in the process are taken by additional measures against undetectable errors be additionally secured.

Claims (8)

  1. Method for operating a fail safe multicomputer system (e.g. RS1) with some not fail safe, input/output modules (E/A11, E/A12), via which the computers of the multicomputer system (RS1) transmit on two-channels (B1, B2) data telegrams from and/or to other computers and/or data receiving, transmitting or processing systems (RS2 to RS6) and via which they receive from there or transmit to there user data, protected by test data, in the form of telegrams which correspond in terms of contents, for these input/output modules, characterised in that,
    - at the latest when the secure multicomputer system (RS1) detects non-identical data in these input/output modules (E/A11, E/A12) transmitted by a different data receiving, transmitting or processing system (e.g. RS2) on both channels (B1, B2), said multicomputer system (RS1) subjects the data stored there to a redundancy test and determines therefrom the input/output module (e.g. E/A11) with the errored data, and the associated channel (B1),
    - the secure multicomputer system (RS1) then inhibits the further evaluation or addressing of data via this input/output module (E/A11) and the associated channel (B1) in communication with the system (RS2) from which the errored data originate,
    - the secure multicomputer system (RS1) informs the system (RS2), from which it has received the data detected as errored, of this measure and causes the said system to stop the further evaluation or addressing of data by means of that input/output module (E/A21) from which the data detected as errored originate, and the associated channel (B1) in communication with it (RS1),
    - the secure multicomputer system (RS1) causes the said system (RS2) in communication with it to transmit long telegrams, in the form of two telegrams which correspond in terms of contents, to one of the input/output modules (E/A22) which is still operated by it and connected to the other channel (B2), and which causes the redundancy testing of the long telegrams originating from there,
    - the secure multicomputer system (RS1) also combines, even in communication with the respective system (RS2), the data to be transmitted to form long telegrams, transmits the said long telegrams to this system (RS2) via one of the input/output modules (E/A12) still operated by it and connected to the other channel (B2) and causes the said system to perform redundancy testing of the long telegrams,
    - in the event of non-observance of the redundancy conditions by the data stored in one of the input/output modules (E/A12) still operated by it, the secure multicomputer system (RS1) also stops the further evaluation or addressing of the data stored there, and in that the secure multicomputer system (RS1) stops the further evaluation and addressing of the data stored in one of the input/output modules (E/A12) still operated by it if, since the first detection of errored data in one of its input/output modules (E/A11), a defined maximum period of time has elapsed without the fault in the input/output module or modules detected as defective or on the affected transmission channel (B1) having been eliminated.
  2. Method for operating a multicomputer system (e.g. RS1), which is secure in terms of signalling technology, having at least two input/output modules (E/A11, E/A12), which are not secure in terms of signalling technology, via which the computers of the multicomputer system transmit on two channels (B1, B2) data telegrams from and/or to other computers and/or other data receiving, transmitting or processing systems (RS2 to RS6) and via which they receive from there or transmit to there user data, protected by test data, in the form of telegrams which correspond in terms of contents, for these input/output modules, characterised in that
    - at the latest when non-identical data are detected in the said input/output modules (E/A11, E/A12), the secure multicomputer system (e.g. RS1) subjects the data stored there to a redundancy test and determines therefrom the input/output module (e.g. E/A11) with the errored data, and the associated channel (B1),
    - the secure multicomputer system (RS1) then inhibits the further evaluation or addressing of data via this input/output module (E/A11) and the associated channel (B1),
    - the secure multicomputer system then causes the systems (RS2 to RS6) communicating with it to transmit long telegrams, in the form of two telegrams which correspond to one another in terms of contents, to one of the input/output modules (E/A12) still operated by it and connected to the other channel (B2), and causes it to perform the redundancy monitoring of long telegrams originating from there,
    - the secure multicomputer system (RS1) also combines, even in communication with the other systems (RS2 to RS6), the data to be transmitted to form long telegrams, transmits the said long telegrams to the other systems (RS2 to RS6) via one of the input/output modules (E/A12) still operated by it and connected to the other channel (B2) and causes the said systems to perform redundancy testing of the long telegrams,
    - in the event of non-observance of the redundancy conditions by the data stored in one of the input/output modules (E/A12) still operated by it, the secure multicomputer system (RS1) also stops the further evaluation or addressing of the data stored there.
    - and the secure multicomputer system (RS1) stops the further evaluation and addressing of the data stored in one of the input/output modules (E/A12) still operated by it if, since the first detection of errored data in one of its input/output modules (E/A11), a defined maximum period of time has elapsed without the fault in the input/output module or modules detected as defective or on the affected transmission channel having been eliminated.
  3. Method according to Claim 1 or 2, characterised in that, when non-identical data are detected in its input/output modules (E/A11, E/A12) supplied with data which are identical in terms of contents and/or when errors are detected in the long telegrams before the permanent inhibiting of the evaluation of the data stored in the respectively affected input/output modules, the secure multicomputer system (RS1) requests at least one renewed data transmission.
  4. Method according to Claim 1, 2 or 3, characterised in that, when non-identical data are detected in the input/output modules (E/A11, E/A12) occupied by the telegrams which are identical in terms of contents, the secure multicomputer system (RS1) stops the further evaluation or the addressing of data out of the input/output modules in each case involved in the comparison, if the redundancy tests carried out by it for the data in these input/output modules lead to the result that the transmitted data satisfy the test conditions.
  5. Method according to Claim 2 or 3, characterised in that when errored data are detected in one of its input/output modules (E/A11), the multicomputer system (RS1) informs at least the data source (RS2), from which the data stored in errored fashion originate, from which of the input/output modules (E/A21) of said data source these data originate, and in that the respective data source (RS2) subsequently stops the further transmission of data via this input/output module (E/A21) and brings about the transmission of long telegrams via one of its input/output modules (E/A22), still operated then, on another channel (B2), or also itself tests the observance of the prescribed redundancy conditions on the reception of the data via this input/output module.
  6. Method according to Claim 1, 2 or 3, characterized in that the maximum period of time for the continuation of operations of the secure multicomputer system (RS1) is made dependent, from the first detection of errored data in one of its input/output modules (E/A11), on the mean time between failures to be expected for the input/output modules still operated and on the data redundancy selected after the occurrence of the fault.
  7. Method according to Claim 1, 2 or 3, characterized in that the data in the two telegrams of usual length which in each case form the long telegrams are represented in inverted form.
  8. Device for carrying out the method according to at least one of Claims 1 to 7, characterized in that a two-channel bus system (B1, B2) is provided for the data transmission, and in that the input/output modules (E/A1, E/A2), provided for receiving data telegrams which are in each case identical in terms of contents, of the secure multicomputer system (R1, R2) are connected to different channels (B1 or B2) of the bus system, and in that the secure multicomputer system (R1, R2) has a timing element (T1, T2) which is secure in terms of signalling technology and can be set when the evaluation of data stored in an input/output module is inhibited, stops the evaluation of the long telegrams after expiry of the switching time impressed in it and stops the evaluation of the data of a long telegram when transmission or storage errors are detected in the said telegram.
EP85113416A 1984-11-20 1985-10-22 Method for operating a fail-safe multi-computer system with some not fail-safe input/output units Expired - Lifetime EP0182134B1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
AT85113416T ATE72489T1 (en) 1984-11-20 1985-10-22 PROCEDURE FOR OPERATION OF A SIGNAL-TECHNOLOGICALLY SECURE MULTI-COMPUTER SYSTEM WITH SEVERAL SIGNAL-TECHNOLOGICALLY UNSECURE INPUT/OUTPUT MODULES.

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
DE3442418 1984-11-20
DE19843442418 DE3442418A1 (en) 1984-11-20 1984-11-20 METHOD FOR OPERATING A SIGNAL-SECURE MULTIPLE COMPUTER SYSTEM WITH MULTIPLE SIGNAL-NON-SECURE IN / OUTPUT ASSEMBLIES

Publications (3)

Publication Number Publication Date
EP0182134A2 EP0182134A2 (en) 1986-05-28
EP0182134A3 EP0182134A3 (en) 1988-07-20
EP0182134B1 true EP0182134B1 (en) 1992-02-05

Family

ID=6250776

Family Applications (1)

Application Number Title Priority Date Filing Date
EP85113416A Expired - Lifetime EP0182134B1 (en) 1984-11-20 1985-10-22 Method for operating a fail-safe multi-computer system with some not fail-safe input/output units

Country Status (3)

Country Link
EP (1) EP0182134B1 (en)
AT (1) ATE72489T1 (en)
DE (2) DE3442418A1 (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0263973A2 (en) * 1986-10-06 1988-04-20 MAN Technologie Aktiengesellschaft Device and method for serial data exchange between more than two subscribers
EP0742500A2 (en) * 1995-05-11 1996-11-13 Siemens Aktiengesellschaft Fail-safe touch-switch functions and switch functions with error avoidance
US5872827A (en) * 1995-12-01 1999-02-16 Sextant Avionique Method for producing a result and device for the implementation thereof
US6487695B1 (en) 1998-08-01 2002-11-26 Dr. Johannes Heidenhain Gmbh Method for providing fail-safe secure data transmission between a numerical control system and a spatially separate unit

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO1990009633A1 (en) * 1989-02-16 1990-08-23 Siemens Aktiengesellschaft Process for operating a multicomputer system characterized by reliable signal quality
WO1990009632A1 (en) * 1989-02-16 1990-08-23 Siemens Aktiengesellschaft Process for operating a multicomputer system characterized by reliable signal quality
DE10353210A1 (en) * 2003-11-13 2005-06-16 Siemens Ag Secure entry of input values
CN113011786A (en) * 2021-04-22 2021-06-22 华北电力大学 Reliability evaluation method of intelligent substation secondary protection system based on hardware equipment

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE2609107A1 (en) * 1976-03-05 1977-09-08 Standard Elektrik Lorenz Ag Data checking circuit monitoring redundancy checking circuit - is used to detect component fault by duplicating register and interrogate gate
US4099234A (en) * 1976-11-15 1978-07-04 Honeywell Information Systems Inc. Input/output processing system utilizing locked processors
ZA792482B (en) * 1978-06-10 1980-06-25 Signal Co Ltd Railway control signal dynamic output interlocking systems
GB8414692D0 (en) * 1984-06-08 1984-07-11 Norcros Investments Ltd Safety device

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0263973A2 (en) * 1986-10-06 1988-04-20 MAN Technologie Aktiengesellschaft Device and method for serial data exchange between more than two subscribers
EP0263973A3 (en) * 1986-10-06 1990-04-04 MAN Technologie Aktiengesellschaft Device and method for serial data exchange between more than two subscribers
EP0742500A2 (en) * 1995-05-11 1996-11-13 Siemens Aktiengesellschaft Fail-safe touch-switch functions and switch functions with error avoidance
EP0742500A3 (en) * 1995-05-11 1998-04-22 Siemens Aktiengesellschaft Fail-safe touch-switch functions and switch functions with error avoidance
US5872827A (en) * 1995-12-01 1999-02-16 Sextant Avionique Method for producing a result and device for the implementation thereof
US6487695B1 (en) 1998-08-01 2002-11-26 Dr. Johannes Heidenhain Gmbh Method for providing fail-safe secure data transmission between a numerical control system and a spatially separate unit

Also Published As

Publication number Publication date
DE3585361D1 (en) 1992-03-19
EP0182134A2 (en) 1986-05-28
DE3442418A1 (en) 1986-05-22
EP0182134A3 (en) 1988-07-20
ATE72489T1 (en) 1992-02-15

Similar Documents

Publication Publication Date Title
EP1352326B1 (en) Method and device for monitoring a data processing and transmission
EP1631014B2 (en) Method and device for coupling critical processes to a bus
DE19752792B4 (en) Device for self-diagnosis of essentially sporadic errors in serial transmission systems
DE2202231A1 (en) PROCESSING SYSTEM WITH TRIPLE SYSTEM UNITS
EP0182134B1 (en) Method for operating a fail-safe multi-computer system with some not fail-safe input/output units
DE2701925C3 (en) Vehicle control with two on-board computers
EP1024639A1 (en) Method and device for determining of a data transfer reliability
EP1985070B1 (en) Method and apparatus for bus coupling of safety-relevant processes
DE10232272B4 (en) Method for operating a device for a security system
DE3238826A1 (en) CONTROL UNIT FOR THE INPUT / OUTPUT INTERFACE CIRCUITS OF A COMPUTER
EP1469627B1 (en) Method for secure data transfer
EP0410270B1 (en) Method for operating a secure signal transmission interface
EP1640869A2 (en) Method for performing a vote on redundant information
DE3332626A1 (en) CIRCUIT ARRANGEMENT FOR DETECTING STATIC AND DYNAMIC ERRORS IN CIRCUIT ASSEMBLIES
EP1596517B1 (en) Method of transmission of redundantly provided data over a single channel
DE3939631C2 (en)
DE3742117C2 (en)
DE4125812C2 (en) Process for secure data transmission
DE3010803C2 (en) Switching device for a three-computer system in railway systems
DE3605359C2 (en) Computer system with several computers
DE3205217C1 (en) Device for protecting the input of a sequence of data messages into a computer
DE102006007844A1 (en) Method for single-channel bus coupling of safety-critical process, involves transmitting release signal from processing channel to intermediate register through processing channel
EP0281890B1 (en) Security circuit device with a plurality of microcomputers processing the same data
DE102021133657A1 (en) Method and system for securing the exchange of data in a network system for industrial controls
DE1932614C3 (en) Test device for determining a multiple selection of processing units in a data processing system

Legal Events

Date Code Title Description
PUAI Public reference made under article 153(3) epc to a published international application that has entered the european phase

Free format text: ORIGINAL CODE: 0009012

AK Designated contracting states

Kind code of ref document: A2

Designated state(s): AT CH DE LI NL

PUAL Search report despatched

Free format text: ORIGINAL CODE: 0009013

AK Designated contracting states

Kind code of ref document: A3

Designated state(s): AT CH DE LI NL

17P Request for examination filed

Effective date: 19880809

17Q First examination report despatched

Effective date: 19900904

GRAA (expected) grant

Free format text: ORIGINAL CODE: 0009210

AK Designated contracting states

Kind code of ref document: B1

Designated state(s): AT CH DE LI NL

REF Corresponds to:

Ref document number: 72489

Country of ref document: AT

Date of ref document: 19920215

Kind code of ref document: T

REF Corresponds to:

Ref document number: 3585361

Country of ref document: DE

Date of ref document: 19920319

PLBE No opposition filed within time limit

Free format text: ORIGINAL CODE: 0009261

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: NO OPPOSITION FILED WITHIN TIME LIMIT

26N No opposition filed
PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: DE

Payment date: 19981217

Year of fee payment: 14

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: CH

Payment date: 19990119

Year of fee payment: 14

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: AT

Payment date: 19990922

Year of fee payment: 15

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: NL

Payment date: 19991027

Year of fee payment: 15

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: LI

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 19991031

Ref country code: CH

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 19991031

REG Reference to a national code

Ref country code: CH

Ref legal event code: PL

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: DE

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20000801

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: AT

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20001022

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: NL

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20010501

NLV4 Nl: lapsed or anulled due to non-payment of the annual fee

Effective date: 20010501